Replication package: The Scope 1 Bias in Industrial Emission Decomposition — Extended LMDI Scope 1+2 engine, Poland 2008–2022

Replication package for the manuscript "The Scope 1 Bias in Industrial Emission Decomposition: Extended LMDI Evidence from Six European Economies" (Dominiak & Rusowicz; submitted to Journal of Industrial Ecology, September 2026).

Contents: code/ — lmdi_engine_v6.py (two-part LMDI-I Scope 1+2 decomposition, eq. 5–13; produces Tables 1–2 and Online Resource 1 Tables A.1, A.2, S1), lmdi_v6_uncertainty_figs.py (Monte Carlo on the grid emission factor, N = 10,000, three uncertainty specifications; parametric bootstrap, N = 5,000; Table 3, Table S3, Figs. 3–4, Figs. S1, S3), make_figures_from_tables_v6.py (Figs. 2, 5, S2, S4), qa_gate.py (138 automated equality checks between the manuscript tables and the engine output), plus the manuscript and Online Resource 1 in Markdown; data/processed/ — Master_Data_PhD.xlsx (Eurostat nrg_bal_c, env_ac_ainah_r2, nama_10_a64 for Poland, Germany, Czechia, Spain) and EF_grid_PL_2005_2022.csv (Polish grid emission factor: 2005–2013 two-source triangulation as published in Dominiak & Rusowicz 2026, MethodsX 17, 104103; 2014–2022 KOBiZE end-user factor); output/ — result workbooks and all figures. Specification: Scope 1 = Eurostat Air Emissions Accounts (combustion plus process, residence principle); Scope 2 = electricity consumption × grid emission factor; logarithmic-mean weights, year-on-year chaining, residual < 10⁻¹³ Mt. Information cutoff of all inputs: Eurostat and KOBiZE releases as of June 2026. Python ≥ 3.10; requirements.txt included. Code: MIT; data under the terms of the original providers.